Basically, Image Upscaler draws lacking pixels on the place of large ones. It was built with the help of Artificial Intelligence.
First, GAN Generative Adversarial Networks (GAN) has been trained in a tremendous photo library. It helps the app to understand how the land, buildings, etc should look like. Now, when you upload the picture, Image Upscaler scans it, understands what the object is, and then draws the rest of the pixels. As a result, you get a resized photo of high quality.